Skip to content

Pull requests: NVIDIA/Megatron-LM

New pull request New
157 Open 363 Closed
157 Open 363 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

docs: improve code comments module: documentation
#1712 opened Jul 26, 2025 by lorinlee Loading…
Fix a typo on README git checkout module: documentation
#1705 opened Jul 24, 2025 by GindaChen Loading…
BugFix: FP8 Communication Mismatch with --first-last-layers-bf16 in tp-comm-overlap bug Something isn't working module: transformer engine
#1703 opened Jul 24, 2025 by xiaomin-D Loading…
1
Add Support for Packed Sequence Format in GPT Training module: training
#1696 opened Jul 17, 2025 by sbhavani • Draft
Fix the typo in readme module: documentation
#1695 opened Jul 17, 2025 by DNXie Loading…
1
Align import to existing module module: data pipeline
#1692 opened Jul 15, 2025 by AlexanderLavelle Loading…
fix(mtp logging): Correctly accumulate MTP loss for logging when log_interval > 1 module: moe
#1684 opened Jul 11, 2025 by Luowaterbi Loading…
1
Update pretrain_mamba.py bug Something isn't working module: documentation
#1682 opened Jul 11, 2025 by vignesh1507 Loading…
[feat, moe] Add support for global aux loss module: moe
#1681 opened Jul 11, 2025 by Victarry Loading…
1
Issue 1672 fix: initializing the current pointed with int64 to avoid … bug Something isn't working
#1673 opened Jul 7, 2025 by sharanmayank Loading…
1
Support 1f1b a2a overlap module: distributed
#1671 opened Jul 7, 2025 by lhb8125 Loading…
1
moe: remove unused variable scale_up module: moe
#1670 opened Jul 6, 2025 by WineChord Loading…
Speed up model parallel initialization module: distributed
#1662 opened Jul 2, 2025 by alexqdh Loading…
Update README.md module: documentation
#1660 opened Jul 2, 2025 by 21jun Loading…
1
bug fixed: wandb artifact requires the tracker file module: debugging
#1654 opened Jun 27, 2025 by yezhengmao1 Loading…
Apply roll operation to position_ids in MTP module: moe
#1651 opened Jun 26, 2025 by iansheng Loading…
@iansheng @BoxiangW
fix twice allgather in moe distrib optimizer module: moe
#1645 opened Jun 23, 2025 by irobot2013-why Loading…
1
Fix log-timer-to-tensorboard on logging module: debugging
#1631 opened Jun 13, 2025 by wplf Loading…
Fix typos: vritual → virtual and decoeder → decoder module: documentation
#1626 opened Jun 11, 2025 by EricLabile Loading…
Fix: Apply q_layernorm consistently in MLA LoRA path module: fine-tuning
#1624 opened Jun 11, 2025 by Flink-ddd Loading…
1
fix: when using moe parallel folding feature and set etp > 1 && ep == 1, the grad sync is incorrect and the loss curve is bad bug Something isn't working module: moe
#1622 opened Jun 10, 2025 by Louis-J Loading…
Set weights_only=False in optimizer module: optimizer
#1618 opened Jun 9, 2025 by zhic-mt Loading…
use a cpu set to cache cuda tensor finished_request_ids module: inference
#1610 opened Jun 5, 2025 by ladyrick Loading…
Add DistTrain, Allow Encoder to Have Different DP Size module: multimodal
#1605 opened May 30, 2025 by zidanehuang001 Loading…
add node_rank argument for example scripts module: training
#1604 opened May 30, 2025 by xylllllllll Loading…
ProTip! Exclude everything labeled bug with -label:bug.